A mobile restaurant recommendation system - application of data mining techniques in m-commerce

Abstract

Using mobile phones to search restaurants in the street is an inconvenient and annoying task. Large amount of results and options are frustrating to users. Recommender systems can help users concentrate on the restaurants that the users may visit. They observe users’ past behavior in order to predict their interested restaurants. One of commonly used technologies is collaborative filtering. It summarizes all users’ activities so that the unknown interests of a particular user can be estimated by using the information from other users.

This project aims to design and develop a restaurant recommender system targeting to mobile phones. User experiences including control simplicity, flexibility and coverage of results, and satisfactory result are the main concerns of the system. This project presents a model combined with neighboring collaborative filtering and Bayesian networks. They are enhanced by using contextual information to extend the variety of prediction sources. With this system, the amount of input requests and irrelevant information would be minimized.